1. Statement of the Technical Field
The present invention relates to the field of computer speech technology and more particularly to the authentication and verification of the identity of users using one or more biometric modalities.
2. Description of the Related Art
Speech based voice services are fast replacing existing touch-tone based interactive voice services. As speech technology matures, it will enable a host of new voice-based services. These services will be used across a wide array of businesses offering numerous products and services through numerous points of contact with the customer. Each time a customer wishes to access any of the voice based services, he or she will generally have to interact with a voice application serving as the front end of the service in question.
Currently, when a voice service is used via telephone there is typically a need to authenticate the caller in order to service that caller. In current systems, the caller generally uses the touch-tone keypad to enter identifying codes and/or PINs. Alternatively, the caller may be able to take advantage speech recognition technology in the voice application to speak commands and codes into the telephonic device communicating with the voice application. With advent of speaker verification technology it is possible to identify a caller based on their voice. However, this technology requires a caller to enroll their voice prior to the voice being used for identification and authentication. As a user uses more and more voice based services, he or she would have to go through the same enrollment process with each new voice application, or would have to settle for more traditional ID and/or PIN methods for authentication. This enrollment process would be cumbersome and subject to error, and may discourage users from using and accessing voice based services.
And while voice remains one of the easiest forms of ubiquitous access, multi-modal handheld devices like deskphones, cellphones, and PDAs are broadening the means for voice access to the various voice applications providing voice services. These multi-modal devices offer not only voice-based means of communication, but also offer other forms of data input which may be used for authentication and verification, such as pen-based input, camera or imaging devices, and scanners, all of which may be used to identify a user and communicate with a voice application.
It would be useful therefore, to provide a method and system that served users across a network of voice applications, such that users would only have to enroll once for using any of the voice applications, and could use each voice service through a variety of authentication means enabled by the technologies incorporated into the various multi-modal devices available to the user.